missing figs, tweaks to text

episodes/02-sources.md

@@ -109,12 +109,24 @@ Scale factors and their uncertainties for many jet identification algorithms in
 * Double-b identification for Higgs bosons: http://cds.cern.ch/record/2866276?ln=en
 * Heavy particle identification (H, W, t): https://arxiv.org/abs/2004.08262
 
+### Analysis method choices
+
+Uncertainties for corrections to physics objects are often provided by CMS, at least for any required corrections, and for the most popular identification algorithms. But uncertainties also arise from analysis strategy choices. There is no way to write an exhaustive list of these uncertainties! But as you build an analysis, consider:
+
+ * Have I calculated any custom corrections for physics objects?
+ * Have I performed any fits to model background and/or signal?
+ * Have I computed any "transfer factors" to model background in a certain region using information from other regions?
+ * Have I designed a validation procedure for my background modeling strategy?
+
+All of these methods, and surely others, will need to be evaluated for uncertainties that should be propagated to your final observable. Uncertainties in fit parameters, statistical uncertainties in various regions, any deviation from expectations in a validation procedure -- all are examples of analysis method uncertainties you might find in CMS papers and consider in your work.
+
 ::::::::::::: keypoints
 
 - Data in CMS plots carry error bars that serve as a "best estimate" of the variance of the Poisson distribution that governs the expected observations for each observable.
 - Systematic uncertainties are assumed to follow Gaussian or log-normal probability distributions.
 - Collision-based uncertainties come from the luminosity and pileup calculations.
 - Detector-based uncertainties come from corrections to the energy response for different physics objects.
 - Scale factor-based uncertainties come from the calculation methods used to measure efficiencies for various reconstruction, identification, isolation, and trigger algorithms.
+- Analysis methods also bring in uncertainties that need to be determined by the analysis team.
 
 :::::::::::::

episodes/03-corrections.md

@@ -31,7 +31,7 @@ Open this website in your browser! https://opendata.cern.ch/eos/opendata/cms/cor
 
 CMS provides all of the corrections that are "required" for sensible analysis results (in most cases), and also provides scale factors for some of the most common identification and isolation algorithms. Prior to 2016, these corrections were shared with analysts as histograms, via PDF files, via TWiki pages, or other forms, and each group calculating a correction could choose the format. For Run 2 data, a common JSON format has been introduced that simplifies and unifies the process of accessing correction values and their uncertainties.
 
-## Exploring the json format
+## Exploring the correction website
 
 When you click on the website link above, you should see the following page:
 
@@ -52,6 +52,9 @@ The columns describe:
    * LHC Run: Run 2! If you revisit this site in 2028 or 2029 you might begin to see some "Run 3" corrections.
  * The files available with corrections. These are clickable links that take you to a summary page for the corrections contained within that file.
 
+
+## Correction file structure
+
 Click on a correction file to explore it!
 
 One of the simplest examples is the [electron correction file](https://opendata.cern.ch/eos/opendata/cms/corrections/jsonpog-integration-summary/summaries/EGM_2016postVFP_UL_electron.html). When clicking on that link, you see a description of the correction contained in the file:
@@ -103,6 +106,8 @@ Table: high-level summary of which types of corrections live in which JSON files
 | Scale factors | Taus | tau |
 | Scale factors | Jets | btagging, ctagging, subjet_btagging, jmar |
 
+## Software for accessing corrections
+
 :::::::::: challenge
 
 ## Install correctionlib